Sebastian Dullien is Senior Policy Fellow at the European Council on Foreign Relations.
When it comes to the internet, the European Union lags well behind the United States. Among the top five dominant internet companies globally, not a single one is European. The so-called Big Five globally (Apple, Google, Facebook, Amazon, and Microsoft) are all American and rake in huge profits based on (partial) monopolies. Worse, in a number of key technologies, Europe probably no longer has the capability to develop this kind of business. For the past decade, the US has attracted the most gifted engineers from Europe, leaving the EU with a relative lack of skill.
The fact that five huge quasi-monopolists are all based in the US matters. We know from economic geography that “clusters” play a very important role in the formation of new firms. Companies are founded where the necessary skills can be found. For core internet firms, this is not in Europe.
Moreover, when a company is sitting on a pile of cash (as Google or Apple is), this company’s chances of remaining ahead increase. All of the Big Five can easily outspend new entrants when it comes to research and development. Google has a competitive edge in running its data centres. After years of experience, it runs them much more efficiently than any newcomer could, and hence can offer cheaper services. To replicate the hardware alone in Google’s data centres, an estimated investment of $30 billion would be necessary – and this does not include the research for creating a match for Google’s superior algorithms and technology. (Just as a comparison: SAP, the biggest EU IT technology firm, has an annual investment budget of about €600 million.) Finally, the Big Five have so much cash that they can (and will) buy any European start-up that really comes up with the next big idea.
All this matters for the EU. The largest share of employment of the Big Five is in the US. The largest share of their wage bill is paid there. The US can use the capabilities of its IT sector for its military developments. And if the Big Five can be induced to pay their fair share of taxes, they will most likely do so in the US.
So, it would clearly be beneficial if Europe had its own digital champions.
Unfortunately, figuring out how to build a digital champion is far more difficult. Creating a single digital market or improving access to venture capital might be good ideas, but they will not create champions.
A single digital market makes doing business easier not only for European companies, but also for the (superior) US digital firms. Access to venture capital might create more start-ups in the EU. But given that the most compelling business model for start-ups is to become attractive enough to be bought up by one of the large US giants, this is also not a sustainable path to competitive digital capabilities.
The best option is to design a European industrial policy for the sector. Of course, such policies are limited by the rules of international agreements such as World Trade Organization (WTO) treaties. However, the agreements leave a number of options for cleverly designed policies.
One possibility would be to use EU data protection standards as an instrument for industrial policy. For example, the EU could legislate that companies should be fined €500,000 for each case in which they hand over personal data of an EU citizen to foreign law enforcement or secret services without a valid EU court order. Given the US rules on how internet companies have to cooperate with US agencies, this would make any kind of engagement by US companies in the EU highly risky and would most likely lead to a withdrawal of a number of US companies from the EU market.
Given the size of the EU market, such a move would create space for home-grown alternatives to Google, Facebook, and Amazon (potentially founded by current employees of US companies with EU passports). Even so, there would be significant downsides: at least for an interim period, European IT customers (including businesses relying on services such as cloud computing) would be faced with internet services of less quality at a higher price. EU companies that rely on these basic services would be hindered, not promoted.
While there are many cases in economic history where industrial policies have worked, there are also many cases where such policies have failed. On the positive side, South Korea comes to mind. It decided (as a very underdeveloped country) to move into the steel and the shipbuilding sectors. It now has the world’s second-largest shipbuilder and the world’s sixth-largest steel producer. China and Japan used industrial policy successfully to support everything from light industries to car manufacturers. Arguably, China’s internet giants Baidou (a search engine), Alibaba (a retailer plus cloud provider), and Tencent (a social network) could only have grown to their current size because China’s “Great Firewall” kept Google, Amazon, and Facebook out of the Chinese markets.
More problematic are attempts such as that of Brazil to establish its own personal computer industry or Malaysia’s to create a competitive national car industry. And mixed cases include that of Airbus, which on the one hand helped to break Boeing’s dominance, but on the other is not a poster child for efficient management.
These risks, costs, and benefits need to be carefully discussed and evaluated when deciding whether the EU should embark on an explicit industrial policy to grow digital champions. But one thing is clear: without government intervention in the market, there is very little chance that Europe will catch up to the US in the core internet business.
edited by Mark Leonard
Stefan Soesanto & Fosca D'Incau